STA 9750 Mini-Project #01: Gourmet Cheeseburgers Across the Globe

Exploring the Most Popular Programming on Netflix

Author

DavidZhai

Task 1: Data Acquisition

Show code

if(!dir.exists(file.path("data", "mp01"))){
    dir.create(file.path("data", "mp01"), showWarnings=FALSE, recursive=TRUE)
}

GLOBAL_TOP_10_FILENAME <- file.path("data", "mp01", "global_top10_alltime.csv")

if(!file.exists(GLOBAL_TOP_10_FILENAME)){
    download.file("https://www.netflix.com/tudum/top10/data/all-weeks-global.tsv", 
                  destfile=GLOBAL_TOP_10_FILENAME)
}

COUNTRY_TOP_10_FILENAME <- file.path("data", "mp01", "country_top10_alltime.csv")

if(!file.exists(COUNTRY_TOP_10_FILENAME)){
    download.file("https://www.netflix.com/tudum/top10/data/all-weeks-countries.tsv", 
                  destfile=COUNTRY_TOP_10_FILENAME)
}

Task:2 Data Cleaning

Show code

if(!require("tidyverse")) install.packages("tidyverse")

Loading required package: tidyverse

── Attaching core tidyverse packages ──────────────────────── tidyverse 2.0.0 ──
✔ dplyr     1.1.4     ✔ readr     2.1.5
✔ forcats   1.0.0     ✔ stringr   1.5.1
✔ ggplot2   3.5.0     ✔ tibble    3.2.1
✔ lubridate 1.9.3     ✔ tidyr     1.3.1
✔ purrr     1.0.2     
── Conflicts ────────────────────────────────────────── tidyverse_conflicts() ──
✖ dplyr::filter() masks stats::filter()
✖ dplyr::lag()    masks stats::lag()
ℹ Use the conflicted package (<http://conflicted.r-lib.org/>) to force all conflicts to become errors

Show code

library(readr)
library(dplyr)

GLOBAL_TOP_10 <- read_tsv(GLOBAL_TOP_10_FILENAME,na="N/A")

Warning: One or more parsing issues, call `problems()` on your data frame for details,
e.g.:
  dat <- vroom(...)
  problems(dat)

Rows: 8840 Columns: 9
── Column specification ────────────────────────────────────────────────────────
Delimiter: "\t"
chr  (3): category, show_title, season_title
dbl  (5): weekly_rank, weekly_hours_viewed, runtime, weekly_views, cumulativ...
date (1): week

ℹ Use `spec()` to retrieve the full column specification for this data.
ℹ Specify the column types or set `show_col_types = FALSE` to quiet this message.

Show code

#str(GLOBAL_TOP_10)
#glimpse(GLOBAL_TOP_10)

COUNTRY_TOP_10<- read_tsv(COUNTRY_TOP_10_FILENAME,na="N/A")

Rows: 411760 Columns: 8
── Column specification ────────────────────────────────────────────────────────
Delimiter: "\t"
chr  (5): country_name, country_iso2, category, show_title, season_title
dbl  (2): weekly_rank, cumulative_weeks_in_top_10
date (1): week

ℹ Use `spec()` to retrieve the full column specification for this data.
ℹ Specify the column types or set `show_col_types = FALSE` to quiet this message.

Show code

#str(GLOBAL_TOP_10)
#glimpse(GLOBAL_TOP_10)

task 2: data cleaning

Show code

#GLOBAL_TOP_10 <- GLOBAL_TOP_10 |>
 # mutate(season_title = if_else(season_title == "N/A",NA_character_, season_title))

#COUNTRY_TOP_10<- COUNTRY_TOP_10 |>
 # mutate(season_title = if_else(season_title == "N/A",NA_character_, season_title))

task 3:Data Import

#making import of data start off with "N/A" as native NA
#original:
#GLOBAL_TOP_10 <- read_tsv(GLOBAL_TOP_10_FILENAME)
#editded
#GLOBAL_TOP_10 <- read_tsv(GLOBAL_TOP_10_FILENAME,na="N/A")

initial data exploration

Show code

library(DT)
GLOBAL_TOP_10 |> 
    head(n=20) |>
    datatable(options=list(searching=FALSE, info=FALSE))

Show code

COUNTRY_TOP_10 |> 
    head(n=20) |>
    datatable(options=list(searching=FALSE, info=FALSE))

get rid of “_”

Show code

library(stringr)
format_titles <- function(df){
    colnames(df) <- str_replace_all(colnames(df), "_", " ") |> str_to_title()
    df
}
GLOBAL_TOP_10 |> 
    format_titles() |>
    head(n=20) |>
    datatable(options=list(searching=FALSE, info=FALSE)) |>
    formatRound(c('Weekly Hours Viewed', 'Weekly Views'))

dropping season title

Show code

GLOBAL_TOP_10 |> 
    select(-season_title) |>
    format_titles() |>
    head(n=20) |>
    datatable(options=list(searching=FALSE, info=FALSE)) |>
    formatRound(c('Weekly Hours Viewed', 'Weekly Views'))

converting hours to min

Show code

GLOBAL_TOP_10 |> 
    mutate(`runtime_(minutes)` = round(60 * runtime)) |>
    select(-season_title, 
           -runtime) |>
    format_titles() |>
    head(n=20) |>
    datatable(options=list(searching=FALSE, info=FALSE)) |>
    formatRound(c('Weekly Hours Viewed', 'Weekly Views'))

Task 4 Exploratory Questions

1. How many different countries does Netflix operate in? (You can use the viewing history as a proxy for countries in which Netflix operates.

Show code

 country <- COUNTRY_TOP_10 |> 
  distinct(country_name) |> 
  nrow()

COUNTRY_TOP_10 |> 
  distinct(country_name) |> 
  format_titles() |>
  datatable(options = list(searching = FALSE, info = FALSE))

Netflix operates in 94 different countries

2. Which non-English-language film has spent the most cumulative weeks in the global top 10? How many weeks did it spend?

Show code

  library(stringr)
  library(dplyr)
  library(DT)
  GLOBAL_TOP_10 |>
  select(category,show_title,season_title,cumulative_weeks_in_top_10) |>  
  filter(category=="Films (Non-English)")|>
  filter(cumulative_weeks_in_top_10 == max(cumulative_weeks_in_top_10)) |>
  datatable(options = list(searching = FALSE, info = FALSE))

The non-English film that has spent the most cumulative week in top 10 is All Quiet on the Western Front, with 23 weeks in top 10

3. What is the longest film (English or non-English) to have ever appeared in the Netflix global Top 10? How long is it in minutes

Show code

  library(stringr)
  library(dplyr)
  library(DT)
GLOBAL_TOP_10 |>
mutate(`runtime_(minutes)` = round(60 * runtime)) |>
filter(category=='Films (English)'| category=='Films (Non-English)' ) |>  
filter(`runtime_(minutes)`==max(`runtime_(minutes)`,na.rm = TRUE)) |>
format_titles() |>
datatable(options = list(searching = FALSE, info = FALSE)) |>
formatRound(c('Runtime (Minutes)' ))

Longest film in English or Non-English to have every appeared in Netflix global top 10 is Pushpa 2: The Rule (Reloaded Version) with a runtime of 224 min

4. For each of the four categories, what program has the most total hours of global viewership?

Show code

  library(stringr)
  library(dplyr)
  library(DT)
GLOBAL_TOP_10 |>
select(category, show_title, season_title, weekly_hours_viewed) |>
group_by(category)|>
filter(weekly_hours_viewed==max(weekly_hours_viewed)) |>
format_titles() |>
datatable(options = list(searching = FALSE, info = FALSE))

These are the 4 programs at the top of their respective categories

5. Which TV show had the longest run in a country’s Top 10? How long was this run and in what country did it occur?

Show code

  COUNTRY_TOP_10 |>
  select(country_name,category,show_title,season_title,cumulative_weeks_in_top_10)|>
  filter(cumulative_weeks_in_top_10==max(cumulative_weeks_in_top_10))|>
  filter(category=="TV") |>
  format_titles() |>
  datatable(options = list(searching = FALSE, info = FALSE))

The TV show that had the longest run in a country’s top 10 is named Money Heist. The it stayed in the top 10 for 127 cumulative weeks, and occurred in Pakistan

6. Netflix provides over 200 weeks of service history for all but one country in our data set. Which country is this and when did Netflix cease operations in that country?

Show code

COUNTRY_TOP_10 |>
group_by(country_name)|>
summarize(weekcount=n_distinct(week))|>
filter(weekcount==min(weekcount)) |> 
format_titles() |>
datatable(options = list(searching = FALSE, info = FALSE))

Show code

COUNTRY_TOP_10 |>
group_by(country_name)|>
filter(country_name=="Russia")|>
filter(week==max(week)) |> 
format_titles() |>
datatable(options = list(searching = FALSE, info = FALSE))

The country that has only 35 weeks of having Netflix services is Russia. From what I can observe Netflix has stopped operating in 02/27/2022

7. What is the total viewership of the TV show Squid Game? Note that there are three seasons total and we are looking for the total number of hours watched across all seasons.

Show code

GLOBAL_TOP_10|>
  select(show_title,weekly_views,category)|>
  filter(show_title=="Squid Game")|>
  group_by(show_title,category) |>
  summarize(totalview=sum(weekly_views,na.rm = TRUE))|>
  format_titles() |>
  datatable(options = list(searching = FALSE, info = FALSE))

`summarise()` has grouped output by 'show_title'. You can override using the
`.groups` argument.

Across all seasons, the TV show Squid Game has a total viewership of 396,900,000

8. The movie Red Notice has a runtime of 1 hour and 58 minutes. Approximately how many views did it receive in 2021? Note that Netflix does not provide the weekly_views values that far back in the past, but you can compute it yourself using the total view time and the runtime.

Show code

GLOBAL_TOP_10|>
  mutate(`Hour_viewed_(minutes)` = round(60 * weekly_hours_viewed)) |>
  group_by(show_title)|>
  filter(show_title=="Red Notice")|>
  filter(year(week)==2021)|>
  # stated that runtime of red notice is 1hr and 58min == 118min
  summarise(year_view=round((sum(`Hour_viewed_(minutes)`,na.rm = TRUE)/118)))|>
  format_titles() |>
  datatable(options = list(searching = FALSE, info = FALSE))

Red Notice was viewed approximately 201,732,203 times in 2021

9. How many Films reached Number 1 in the US but did not originally debut there? That is, find films that first appeared on the Top 10 chart at, e.g., Number 4 but then became more popular and eventually hit Number 1? What is the most recent film to pull this off?

Show code

COUNTRY_TOP_10|>
  group_by(show_title)|>
  mutate(weekly_rank,toprank=weekly_rank==1)|>
  mutate(debutTR=(toprank==TRUE)&(week==min(week)))|>
  filter(toprank==TRUE,debutTR==FALSE,country_name=="United States",category=="Films")|>
  format_titles() |>
  datatable(options = list(searching = FALSE, info = FALSE))

130 Films reached Number 1 in the US but did not originally debut there. KPop Demon Hunters is the most recent film to pull this off.

10. Which TV show/season hit the top 10 in the most countries in its debut week? In how many countries did it chart?

Show code

COUNTRY_TOP_10|>
  group_by(show_title)|>
  select(country_name,week,category,weekly_rank,show_title,season_title,cumulative_weeks_in_top_10)|>
  mutate(weekly_rank,top_ten=(weekly_rank>=1 & weekly_rank<=10))|>
  mutate(debut_top_ten=(top_ten==TRUE)&(week==min(week)))|>
  filter(debut_top_ten==TRUE,category=="TV")|>
  summarize(countries_charted = n_distinct(country_name),debut_top_10 = TRUE) |>
  arrange(desc(countries_charted)) |>
  format_titles() |>
  datatable(options = list(searching = FALSE, info = FALSE))

Emily in Paris has charted in most countries around the world in its debut week of 94 different countries

Task 5 :Write a press release promoting season 5 of Stranger Things. Be sure to include a catchy headline.

Netflix will release the fifth and final season of its hit show Stranger Things at the end of 2025. In preparation for the release, prepare a press release highlighting the broad impact of the previous four seasons. Your press release should refer to both the total viewership, the length of popularity (how many weeks in the top 10) and the multinational appeal of previous seasons. You should also compare the impact of Stranger Things to other popular English-language TV shows to give a sense of its success.

Show code

library(DT)
library(tidyverse)
library(dplyr)
library(stringr)
library(lubridate)
library(ggplot2)

GLOBAL_TOP_10|>
  mutate(season_title=ifelse(is.na(season_title)& show_title=="Stranger Things","Stranger Things 1",season_title))|>
  filter(show_title=="Stranger Things")|>
  group_by(season_title)|>
  summarize(
    total_week_in_top10=sum(cumulative_weeks_in_top_10,na.rm=TRUE),
    total_hours_viewed    = sum(weekly_hours_viewed, na.rm = TRUE)
    )|>
  format_titles() |>
  datatable(options = list(searching = FALSE, info = FALSE))

Show code

#cumulative_weeks_in_top_10 graph
GLOBAL_TOP_10 |>
  filter(show_title == "Stranger Things") |>
  mutate(season_title = ifelse(is.na(season_title), "Stranger Things 1", season_title)) |>
  group_by(season_title, week) |>
  ggplot(aes(x = week, y = cumulative_weeks_in_top_10, color = season_title)) +
  geom_line(size = 1) +
  labs(
    title = "cumulative_weeks_in_top_10 of Stranger Things by Season",
    x = "Week",
    y = "cumulative_weeks_in_top_10",
    color = "Season"
  ) +
  theme_minimal()

Warning: Using `size` aesthetic for lines was deprecated in ggplot2 3.4.0.
ℹ Please use `linewidth` instead.

Show code

stranger <-GLOBAL_TOP_10|>
  filter(show_title=="Stranger Things")|>
  group_by(week)|>
  summarise((total_week_viewed=sum(weekly_hours_viewed,na.rm=TRUE)))|>
  arrange(week)
colnames(stranger)<-c("Week","total_hours_viewed_in_a_week")

GLOBAL_TOP_10 |>
  filter(show_title == "Stranger Things") |>
  group_by(week) |>
  arrange(week)|>
  ggplot(aes(x = week, y = weekly_hours_viewed)) +
  geom_line(size = 1) +
  labs(
    title = "weekly_hours_viewed of Stranger Things",
    x = "Week",
    y = "weekly_hours_viewed"
  ) +
  theme_minimal()

Show code

GLOBAL_TOP_10 |>
  filter(show_title == "Stranger Things") |>
  mutate(season_title = ifelse(is.na(season_title), "Stranger Things 1", season_title)) |>
  group_by(season_title, week) |>
  ggplot(aes(x = week, y = weekly_hours_viewed, color = season_title)) +
  geom_line(size = 1) +
  labs(
    title = "weekly_hours_viewed of Stranger Things by Season",
    x = "Week",
    y = "weekly_hours_viewed",
    color = "Season"
  ) +
  theme_minimal()

Show code

#graphs look strange  because recorded data of stranger things n=20

Vecna commands the world’s gaze : Stranger Things Season 5 Prepares to Captivate the Globe

Vecna has the world in his grip, and fans everywhere are powerless to look away. Stranger Things Season 5 arrives at the end of 2025 after Season 4 that dominated the Top 10 across more countries than any previous season, with peak weekly hours viewed skyrocketing as audiences everywhere watched. Season 4’s unprecedented cumulative weeks in the Top 10 proved the series’ lasting global appeal, and as Hawkins faces its final battle, the world prepares for the epic conclusion to a phenomenon that has captured hearts and screens across the globe.
Stanger Things across its four seasons has grossed 2,967,980,000 hours viewed internationally, and spent an astonishing 366 weeks cumulative weeks in Netflix’s top 10 globally. As season 5 approaches, this record setting run emphasizes the investment audiences have in Eleven and her friends in the final chapter of their stories. See you guys at Hawkins !

Task 6: Netflix in India Press Release

Write a press release touting the success of Netflix in India and highlighting recent subscriber growth in the region. Be sure to include a catchy headline.

Show code

library(DT)
library(dplyr)
library(stringr)
library(ggplot2)


COUNTRY_TOP_10 %>%
  filter(country_name == "India") |>
  group_by(week)|>
  summarise(total_cumulative_weeks = sum(cumulative_weeks_in_top_10, na.rm = TRUE)) |>
  arrange(week) |>
  mutate(ma_4week = stats::filter(total_cumulative_weeks, rep(1/4, 4), sides = 1)) |>
  ggplot(aes(x = week)) +
  geom_line(aes(y = total_cumulative_weeks), color = "red", size = 1.2) +
  geom_line(aes(y = ma_4week), color = "blue", size = 1.2) +  # MA line
  labs(
    title = "Netflix Engagement Growth in India: Cumulative Weeks in Top 10 (with 4-week MA)",
    x = "Week",
    y = "Total Cumulative Weeks in Top 10"
  ) +
  theme_minimal()

Warning: Removed 3 rows containing missing values or values outside the scale range
(`geom_line()`).

Show code

#dt for global inner join with country is too big, had to make it dataframe
india_global <-COUNTRY_TOP_10 |>
  filter(country_name == "India") |>
  inner_join(
    GLOBAL_TOP_10,
    by = c("show_title","week"),
    suffix = c("_india", "_global"),
    relationship = "many-to-many"  
  ) |>
  select(
    show_title,
    week,
    category_india,
    category_global,
    weekly_hours_viewed,
    runtime,
    weekly_views
  )
india_global

# A tibble: 2,942 × 7
   show_title      week       category_india category_global weekly_hours_viewed
   <chr>           <date>     <chr>          <chr>                         <dbl>
 1 Saiyaara        2025-09-21 Films          Films (Non-Eng…            11900000
 2 Mahavatar Nars… 2025-09-21 Films          Films (Non-Eng…             5000000
 3 Inspector Zende 2025-09-21 Films          Films (Non-Eng…             2000000
 4 The Wrong Paris 2025-09-21 Films          Films (English)            38900000
 5 The Ba***ds of… 2025-09-21 TV             TV (Non-Englis…            14800000
 6 Bon Appétit, Y… 2025-09-21 TV             TV (Non-Englis…            84700000
 7 Wednesday       2025-09-21 TV             TV (English)               56200000
 8 Wednesday       2025-09-21 TV             TV (English)               14500000
 9 Adolescence     2025-09-21 TV             TV (English)               13700000
10 Wednesday       2025-09-21 TV             TV (English)               56200000
# ℹ 2,932 more rows
# ℹ 2 more variables: runtime <dbl>, weekly_views <dbl>

Show code

india_global_cumulative<- india_global|>
  filter(!is.na(weekly_views))|>
  group_by(week)|>
  summarise(total_cumulative_weekly_views= sum(weekly_views, na.rm = TRUE))|>
  arrange(week)|>
  mutate(cum_views = cumsum(total_cumulative_weekly_views))

  ma_4week2 = stats::filter(india_global_cumulative$total_cumulative_weekly_views, rep(1/4, 4), sides = 1)

ggplot(india_global_cumulative, aes(x = week))+
  geom_line(aes(y = total_cumulative_weekly_views),color="red",size=1.2) +  
  geom_line(aes(y = ma_4week2), color="blue", size=1.2) +
    labs(
    title = "India weekly views by week",
    x="Week",
    y="Hours viewed by week"
  )+
  theme_minimal()

Warning: Removed 3 rows containing missing values or values outside the scale range
(`geom_line()`).

Show code

ggplot(india_global_cumulative, aes(x = week, y = cum_views)) +
  geom_line(color = "darkblue", size = 1.2) +
  labs(
    title = "Cumulative Netflix Views in India",
    x = "Week",
    y = "Cumulative Views"
  ) +
  theme_minimal()

From Mumbai to Millions: Netflix India Surges to New Heights

Analysis of Netflix’s time series reveal total weekly viewing hours surge of over 150% from mid-2023 and the peak acheived in early 2025. This explosive growth in engagement translate to a significant increase in our active and engaged customer base, suggesting the strategic focus on localized content is yielding exponential returns. India is one of the most populous country, and provides an opportunity for Netflix and a chance for Hindi content to flourish.
While tracking the Cumulative Weeks in Top 10, smoothed by a 4-week moving average, shows a clear upward trend since 2022. This shows that the growth of the platforms success in consistently delivering a broad and diverse catalouge that resonates with local audiences.
The long-term growth indicated by the viewing hours moving average is decisively upward. The consistently rising baseline for engagement, paired with increasing cumulative weeks in the Top 10, reflects strong demand from Indian viewers for Netflix’s diverse content.

Task 7: 3rd Press Release

Netflix decodes global viewer: Data reveals UNQUENCHABLE appetite for ROMANCE!… and sequels

By analyzing the words most frequently used in the titles that reached the Top 10 globally, Netflix has isolated three core areas of viewer demand:
1. The enduring power of love: The word “love” ranked #1 with 229 occurrences in top 10 titles, demonstrating a massive year round appetite for romance, rom-coms, and dramas
2. The Power of Sequels: Believe it or not the #2 ranked word is “2” with 208 occurrences. This can confirm that audiences place high value on established franchises. Sequels, second seasons, part 2s, carry recognition from fans from the original content, while carrying significantly less risk for Netlfix.
3. The Quest for Relatability: The word “Life” ranked third with 107 occurrences, signaling a strong demand for character-driven narratives. These titles—spanning biopics and documentaries—appeal to viewers seeking emotional depth and stories that resonate with universal human experiences.
4. Holiday Guarantee: The keyword “Christmas” (85 occurrences) confirms its powerful, cyclical role in content programming. While performance may fluctuate year-to-year, holiday-themed content reliably peaks in engagement near the end of every year, making it a cornerstone of Q4 scheduling
5. Absolute Cinema : The keyword “Movie” (85 occurrences) highlights the continued importance of the film format. Shows the viewers want for a complete, high quality single sitting experience

Show code

library(tidytext)


pop_word <- GLOBAL_TOP_10 |>
  select(show_title) |>
  unnest_tokens(word, show_title) |>
  filter(!word %in% stop_words$word)|>
  count(word, sort = TRUE)
pop_word|>
  format_titles() |>
  datatable(options = list(searching = FALSE, info = FALSE))

Show code

top_10_words <- pop_word |> 
  slice_max(n, n = 10)
ggplot(top_10_words, aes(x = reorder(word, n), y = n, fill = word)) +
  geom_col() +
  coord_flip() +  # horizontal bars make it easier to read
  labs(
    title = "Top 10 Words in Netflix Show Titles",
    x = "Word",
    y = "Frequency"
  ) +
  theme_minimal() +
  theme(legend.position = "none")

Show code

word_data <- GLOBAL_TOP_10 |>
  select(week, show_title) |>
  mutate(title_lower = tolower(show_title)) |>
  unnest_tokens(word, title_lower) |>
  filter(!word %in% stop_words$word)
# Count total word frequency across all weeks
keyword_freq <- word_data |>
  group_by(word)|>
  summarise(total_count = n(), .groups = "drop") |>
  arrange(desc(total_count))
# Take top 5 words based on total count
top_words <- keyword_freq$word[1:5]
# Count frequency per week for those top words
word_weekly <- word_data |>
  filter(word %in% top_words) |>
  group_by (week, word) |>
  summarise(freq = n())

`summarise()` has grouped output by 'week'. You can override using the
`.groups` argument.

Show code

word_weekly <- word_weekly %>%
  mutate(color = case_when(
    word == top_words[1] ~ "red",
    word == top_words[2] ~ "blue",
    word == top_words[3] ~ "green",
    word == top_words[4] ~ "orange",
    word == top_words[5] ~ "purple"
  ))

ggplot(word_weekly, aes(x = week, y = freq, color = color)) +
  geom_line(size = 1) +
  facet_wrap(~ word, scales = "free_y") +
  theme_minimal() +
  theme(axis.text.x = element_text(angle = 45, hjust = 1)) +
  guides(color = "none")